1. In aerobic conditions the isolated perfused liver from well-fed rats rapidly formed lactate from endogenous glycogen until the lactate concentration in the perfusion medium reached about 2mm (i.e. the concentration of lactate in blood in vivo) and then production ceased. Pyruvate was formed in proportion to the lactate, the [lactate]/[pyruvate] ratio remaining between 8 and 15. 2. The addition of 5mm- or 10mm-glucose did not affect lactate production, but 20mm- and 40mm-glucose greatly increased lactate production. This effect of high glucose concentration can be accounted for by the activity of glucokinase. 3. The perfused liver released glucose into the medium until the concentration was about 6mm. When 5mm- or 10mm-glucose was added to the medium much less glucose was released. 4. At high glucose concentrations (40mm) more glucose was taken up than lactate and pyruvate were produced; the excess of glucose was probably converted into glycogen. 5. In anaerobic conditions, livers of well-fed rats produced lactate at relatively high rates (2.5mumol/min per g wet wt.). Glucose was also rapidly released, at an initial rate of 3.2mumol/min per g wet wt. Both lactate and glucose production ceased when the liver glycogen was depleted. 6. Addition of 20mm-glucose increased the rate of anaerobic production of lactate. 7. d-Fructose also increased anaerobic production of lactate. In the presence of 20mm-fructose some glucose was formed anaerobically from fructose. 8. In the perfused liver from starved rats the rate of lactate formation was very low and the increase after addition of glucose and fructose was slight. 9. The glycolytic capacity of the liver from well-fed rats is equivalent to its capacity for fatty acid synthesis and it is pointed out that hepatic glycolysis (producing acetyl-CoA in aerobic conditions) is not primarily an energy-providing process but part of the mechanism converting carbohydrate into fat.